LLM agents often give conflicting answers to the same question, and many deployed agents are variants of one base model, so their errors are correlated. We study truth discovery among such agents: inferring the true answer and each agent's reliability without labels. Classical truth-discovery and crowdsourcing estimato...
Agile Earth observation satellites must coordinate imaging tasks within limited observation windows, while accounting for attitude maneuvering and onboard energy constraints. Competition among observation opportunities makes scheduling performance sensitive to the heuristic rules used for task selection. This letter pr...
Inoculation prompting is a method to reduce undesired behaviours such as reward hacking in frontier models. It works by training a model on harmful examples under a prompt that specifies the undesired behaviour, which reduces the model's tendency to exhibit it unless prompted. However, the inoculation prompt is written...
Real-world problems span diverse domains, including program synthesis, symbolic reasoning, planning, and optimization, and often require fundamentally different solution paradigms. A central challenge for LLM-based reasoning is twofold: identifying an appropriate problem-solving approach and formulating it correctly. I...
Yilun Hao, Ruixiao Yang, Hudson Hilal et al.· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
WASP-Mem (Wiki-based Agent-Synthesized Procedural Memory), a nimble but effective method that distills execution traces and feedback into persistent, interlinked wiki pages, enabling continual agent learning is introduced.
Pei-Zhi Niu, Ge Shi, Vaibhav Gorde et al.· 0 citations
OLIVIA, a protocol for point-in-time inputs with company identities masked, selection on validation data only, an analysis plan frozen by cryptographic hash, and a test period unblinded once through a logged lock is proposed.
A. Dehkordi, Alireza Kargarzadeh, Nariman Khaledian et al.· 0 citations
Language models serve as the reasoning core of agents for chemistry and materials science, where they must interpret molecular structure before planning syntheses or proposing materials. Yet their outputs show only what they say, not what they know. Given a prompt such as ``The primary functional group in 1-Fluorobutan...
T. Nguyen, Anh-Vu Nguyen, O. Issakah et al.· 0 citations
It is argued that compliance survives a change in perceived oversight is a property of the model--value pairing, not of the model alone, and value-dependent: whether compliance survives a change in perceived oversight is a property of the model--value pairing, not of the model alone.
Manodeep Ray, Vivek Srivastava, S. Karande et al.· 0 citations
A vision for a system called TimeLess, designed to reshape the software development process by enabling immediate action during meetings and allowing teams to focus on critical decision-making while AI agents execute development tasks based on the meeting discussions.
Z. Rasheed, M. Sami, Jussi Rasku et al.· arXiv.org· 6 citations
Good knowledge, positive attitude and some sociodemographic factors were significant with exclusive breastfeeding which are needed to improve for promoting exclusive breastfeeding practices and reducing infant mortality and morbidity.
M. Hasan, M. Hassan, M. Khan et al.· Population Medicine· 32 citations· ⚡6
Children suffering from severe acute malnutrition (SAM) have a weakened immune system. The relationship between malnutrition and alterations in the frequency of peripheral blood mononuclear cells (PBMCs) remains unclear. This study investigated the altered immune responses in Bangladeshi children with SAM compared to h...
Zannatun Noor, S. Islam, M. Hasan et al.· Immunobiology· 3 citations
This paper develops two new encoding-decoding constructions that attain the previously established lower bound on the achievable miss-detection probability in the finite-token regime, thereby completely characterizing the optimal multi-bit watermarking performance.
Flint-interactive gives AI agents a language for creating interactive visualizations. Click a chart, filter data, highlight related content, and explore beyond the view the author originally chose. The post Flint-interactive: A language for creating interactive data visualizations appeared first on Microsoft Research.
